382 C. M. Child 
and removed from the aboral end in some cases in both series 
(c, Fig. 31), but with the usual method of experiment extrusion 
occurs chiefly from the oral end. Fig. 31 is of course purely dia- 
grammatic and the different lines serve merely to indicate the effect 
of removal of more or less. 
It is evident, however, that the more complete the extrusion and 
removal at the oral end the greater the distance over which the 
mesenteries are almost wholly removed (compare the lines a and 3, 
Fig. 31). At present I am inclined to believe that these differences 
are factors in the oral-aboral localization of the tentacle groups. 
Apparently in those cases where the degree of injury at the two 
ends is not very different (e. g., after removal along a and c, Fig. 
31) the tentacle groups form on the line of union between oral and 
aboral ends, while in those cases where the mesenteries are much 
more completely cut away from the body-wall at the oral end (e. g., 
by a second operation at b, Fig. 31, the groups are localized on the 
oral side of the line of union. In the first series of cases oral and 
aboral ends are involved in much the same degree in the forma- 
tion of the new tissue at the line of union, while in the second loss 
of the old differentiation and growth occur to a larger extent on 
the oral side of the line of union. Since the tentacle groups are 
localized in the region of most active growth, they appear in the 
first series at the line of union and in the second on the oral side of 
it, 
This is the only factor, if it is actually a factor, of importance 
for the oral-aboral localization of the tentacle groups, which is 
apparent from my observations. Whether this is the chief factor, 
or whether other factors are concerned, must remain for the future 
to determine. ‘The only fact that might indicate that other factors 
play a part in determining this localization is that no cases have 
been observed in which some of the tentacle groups were on the 
line of union and some wholly oral to it: all the groups on each 
ring are apparently in the one place or the other. If the character 
of the operation is the chief factor the possibility of the appearance 
of both tocalizations in a single ring must be admitted, though it 
need not be expected in any given case. Further experiments are 
necessary to determine this point. 
